Contents lists available at ScienceDirect

International Journal of Hygiene andEnvironmental Health

journal homepage: www.elsevier.com/locate/ijheh

Occurrence and speciation of chlorination byproducts in marine waters andsediments of a semi-enclosed bay exposed to industrial chlorinated effluents

Tarek Manasfia,∗, Karine Lebarona, Maxime Verlandea, Julien Dronb, Carine Demelasa,Laurent Vassaloa, Gautier Revenkob, Etienne Quiveta, Jean-Luc Boudennea

a Aix Marseille Univ, CNRS, LCE UMR7376, 13331, Marseille, Franceb Institut Ecocitoyen Pour la Connaissance des Pollutions, Centre de vie la Fossette RD 268, 13270, Fos-sur-Mer, France

A R T I C L E I N F O

Keywords:Chlorination byproductsDisinfection byproductsIndustrial chlorinated effluentsCooling waterMarine sedimentsSeawater pollution

A B S T R A C T

Chlorination of seawater is one of the most effective technologies for industrial biofouling control. However,chlorination leads to the formation of halogenated chlorination byproducts (CBPs) associated with potential risksto environmental and human health. The present study investigated the occurrence and distribution of CBPs inthe Gulf of Fos, a semi-enclosed bay where chlorinated effluents of multiple industrial plants are discharged.Seawater samples (surface and bottom) were collected at 24 sampling stations, with some near industrial outletsand others dispersed throughout the bay. Sediment samples were also collected at 10 sampling stations.Physicochemical parameters including water temperature, pH, salinity, bromide content, and free and totalresidual oxidant were determined. Several chemical classes of CBPs including trihalomethanes, haloacetic acids,haloacetonitriles, trihaloacetaldehydes, and halophenols were analyzed. Bromoform was the most abundant CBPin seawater, and it was detected at most of the sampling stations of the bay with highest concentrations oc-curring near the industrial effluent outlets. Dibromoacetic acid was the second most abundant CBP at most of thesites followed by dibromoacetonitrile. Other detected CBPs included tribromoacetic acid, bromo-chloroacetonitrile, and bromal hydrate. To our knowledge, the concentration of the latter CBP was reported herefor the first time in the context of industrial seawater chlorination. In sediments, two bromine-containing ha-lophenols (2-chloro-4-bromophenol and 2,4,6-tribromophenol) were detected at two sampling stations.Ecotoxicological assays and risk assessment studies based on the detected environmental concentrations arewarranted to elucidate the impacts of marine CBP contamination.

1. Introduction

The use of seawater in industrial cooling or heating is a commonpractice in many parts of the world. One of the primary operationalproblems of using seawater in such processes is biofouling, which re-sults from the growth of microorganisms (biofilms) and macro-organ-isms (e.g., clams) on the surface or inside industrial equipment. Biofilmstend to stick to heat-exchange surfaces, thereby significantly reducingheat-transfer coefficients, while excessive development of macro-or-ganisms can plug heat exchangers. There are several techniques forpreventing both types of biofouling. Chlorination of seawater is amongthe most commonly used antibiofouling treatments (Khalanski andJenner, 2012). Chlorine is added into seawater either in the gaseousform or in the aqueous form of sodium hypochlorite solution, typically

at doses of 0.5–1.5 mg/L (expressed as Cl2) (Allonier et al., 1999a,b; Maet al., 2011; Khalanski and Jenner, 2012). In seawater, chlorine reactswith organic and inorganic compounds leading to the formation ofchlorination byproducts (CBPs) (or disinfection byproducts, DBPs)(Heeb et al., 2014). Several factors including initial chlorine dose,temperature, pH, constitution of seawater and presence of con-taminants (natural or anthropogenic) can influence these reactions,leading to differences in the nature and levels of the formed CBPs(Allonier et al., 1999b; Heeb et al., 2014).

The release of chlorinated seawater into the environment con-stitutes a concern from environmental and human health standpoints.Chemical hazards associated with chlorination of seawater can be di-vided into acute effects from the action of strong oxidants and long-term effects caused by CBPs. While the employed oxidants generally act

https://doi.org/10.1016/j.ijheh.2018.06.008Received 29 January 2018; Received in revised form 1 June 2018; Accepted 28 June 2018

∗ Corresponding author. Aix Marseille Université, 3 Place Victor Hugo, Case 29, 13331, Marseille, France.E-mail addresses: tarek.manasfi@univ-amu.fr (T. Manasfi), karine.lebaron@univ-amu.fr (K. Lebaron), maxime.verlande@univ-amu.fr (M. Verlande),

julien.dron@institut-ecocitoyen.fr (J. Dron), carine.demelas@univ-amu.fr (C. Demelas), laurent.vassalo@univ-amu.fr (L. Vassalo),gautier.revenko@institut-ecocitoyen.fr (G. Revenko), etienne.quivet@univ-amu.fr (E. Quivet), jean-luc.boudenne@univ-amu.fr (J.-L. Boudenne).

International Journal of Hygiene and Environmental Health 222 (2019) 1–8

1438-4639/ © 2018 Elsevier GmbH. All rights reserved.

T

as corrosives on marine fauna and flora, the generated mixtures of CBPspresent far more complex toxicological challenges, to both natural biotaand human health (Werschkun et al., 2014). Although data on theecotoxicity of CBPs are very limited, existing data show that persistentCBPs may induce adverse effects on marine organisms, when presentabove certain concentrations (Taylor, 2006; Deng et al., 2010; Pignataet al., 2012; Khalanski and Jenner, 2012). CBPs have been found toinduce developmental toxicity to marine Polychaete Platynereis dumer-ilii (Yang and Zhang, 2013) and oysters Crassostrea virginica (Stewartet al., 1979). Data about the toxicity (with endpoints relevant to humanhealth) of CBPs in chlorinated drinking water and swimming poolwaters are much more extensive than ecotoxicity data (endpoints re-levant for environmental health). Extensive studies have shown thatCBPs can induce a range of adverse effects including mutagenicity,genotoxicity, and carcinogenicity (Richardson et al., 2007; Manasfiet al., 2017b). Concerns about potential human exposure to CBPs in thecontext of seawater chlorination are not only related to the exposure ofworkers in the vicinity of industrial seawater chlorination sites (Banerjiet al., 2012), but also to the general population as certain persistentCBPs may potentially bioaccumulate in exposed marine organismswhich are consumed as food products (e.g., fish and mussels)(Khalanski and Jenner, 2012; Boudjellaba et al., 2016). The bioaccu-mulation potential of an organic compound depends on its ability toaccumulate in fats of marine organisms. Hence, CBPs having a highoctanol/water partition coefficient (Log Kow) such as 2,4,6-tri-bromophenol (Log Kow= 3.92–4.02) are susceptible to bioaccumula-tion and have been previously detected in marine organisms (Khalanskiand Jenner, 2012; Boudjellaba et al., 2016).

The identification of dominant CBPs discharged into the receivingmarine waters and their concentrations is key for performing environ-mental and human health risk assessment. To date, there have been fewstudies documenting CBPs concentrations in chlorinated industrial ef-fluents and receiving marine waters. Most of the previous studies fo-cused on measuring a limited number of CBPs in cooling effluents ofnuclear and thermal power plants and in a few discharge points in theopen coast (Jenner et al., 1997; Allonier et al., 1999a,b; Khalanski andJenner, 2012). Data about the dissemination of CBP contamination ofseawater and marine sediments exposed to multiple industrial chlori-nated effluents remain very scarce. Boudjellaba et al. (2016) previouslyexplored the occurrence of CBPs in seawater and fish in the Gulf of Fosin Southeastern France, based on sampling campaigns in winter andsummer 2014. The Gulf of Fos is a semi-enclosed bay that favors waterconfinement in some of its back-ends and receives the plumes of thesecond greatest Mediterranean river, namely the Rhône river, amongother freshwater inputs (Ulses et al., 2005). This gulf hosts the largestport of trade in France (Marseille-Fos Port) and a major industrial zonethat includes steel, petrochemical, waste incineration, and cement in-dustries, along with gas and electricity power plants. Despite these in-dustrial activities, amateur and professional fishing are practiced in thebay, and recreational swimming areas are also available at some coastalareas around the bay. In the study of Boudjellaba et al. (2016), CBPclasses which have been previously detected in cooling water effluentswere analyzed. The present study aims at conducting a more compre-hensive survey about the contamination of the Gulf of Fos, by analyzinga broader array of CBPs and by sampling a more extended geographicalzone to further evaluate the CBP contamination of seawater throughoutthe Gulf. Additionally, marine sediments were sampled to evaluate theircontamination by CBPs, for the first time to our knowledge. Here, thesampling campaign was carried out during the spring season, unlike theprevious survey of Boudjellaba et al. (2016). Measuring CBP con-centrations in seawater during different seasons contributes to doc-umenting their levels under different meteorological conditions (espe-cially temperature) and potentially different chlorination practicesperformed by the industries in different meteorological conditions. Forexample, when water temperature is low some industries decreasechlorination as the risk of biofouling decreases at lower temperatures.

Documenting sufficient data about the contamination of seawater byCBPs based on multiple surveys performed during different periods ofthe year would contribute to a better monitoring of the contaminationand to defining trends, which are of high utility for risk assessmentstudies.

To this end, the present study investigated the contamination ofseawater and sediments by CBPs in the Gulf of Fos. Several classes ofhalogenated CBPs including trihalomethanes (THMs), haloacetic acids(HAAs), haloacetonitriles (HANs), trihaloacetaldehydes (THA), andhalophenols (HPs) were analyzed in seawater and marine sedimentsobtained from the vicinity of the industrial effluents and at other sitesthroughout the bay. Global physicochemical parameters such as tem-perature, pH, salinity, bromide concentration, total organic carbon(TOC), total nitrogen (TN), total oxidant residual, and free oxidant re-sidual were also determined.

2. Materials and methods

2.1. Study site

The study investigated the Gulf of Fos (Fig. 1), which hosts nu-merous industrial sites. The Gulf is located in Southeastern France onthe Mediterranean at about 50 km from the city of Marseille. Averagewater depth in the Gulf is about 20m. The Gulf receives severalfreshwater inputs including a main input from the Rhône River, andminor inputs from the Berre Lagoon, irrigation and navigation canals.The region is characterized by frequent and strong north or northwestwinds (around 40% per year) and southeast winds (10–20% per year).The north or northwest wind is most common in winter and spring,although it occurs in all seasons (MeteoFrance). A wind rose showingwind distributions based on normalized data from 2002 to 2011 ispresented in SI (Fig. S2). Various heavy industrial activities are estab-lished around the Gulf including two large liquefied natural gas (LNG)terminals (Fos-Cavaou and Fos-Tonkin designated by sampling stations12x and 8p, respectively). These two LNG terminals discharge chlori-nated waters at a flow of 30,000m3/h and 15,000m3/h followingelectrochlorination or addition of sodium hypochlorite, respectively. Inaddition, there are other power plants which are irregularly active(designated by sampling stations 9x and 11x) which discharge chlori-nated water (by electrochlorination). Metal industry (sampling station10x) and oil refineries (sampling station 13x) also discharge chlorinatedseawater at flows exceeding 10,000m3/h. The outlets of the different

Fig. 1. Overview of the Gulf of Fos and the different sampling stations (from 1cto 24m). Triangles represent industrial effluent discharge points.

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

2

industrial sites in the Gulf of Fos are shown in Fig. 1 (sampling stationnames end with x and are represented in triangles). Maps of the studysite were produced using R (R Core Team, 2015) and Inkscape (2015)software.

2.2. Sample collection

The sampling campaign was performed during spring (24, 25, and26 April 2017). The different sampling stations were disseminatedthroughout the Gulf of Fos and included the main industrial outlets(Fig. 1). Water samples were collected at all the 24 sampling stations atthe surface (depth between 0 and 50 cm) and at 7m depth (or thebottom at the stations where the bottom is at depth < 7m) using a 5-LNiskin bottle (General Oceanics, USA). For the analysis of CBPs, samplealiquots (1 L) were placed in amber glass bottles, ascorbic acid wasrapidly added, and bottles were sealed with PTFE-lined screw caps.These CBP samples were filled without headspace to avoid any loss dueto volatilization. For the analysis of global physicochemical parametersincluding bromide ion concentration, TOC, TN, and determination offree and total chlorine, sample aliquots (1 L) were placed in amber glassbottles with PTFE-lined screw caps to which no ascorbic acid wasadded. Samples were stored at 4 °C away from sunlight and extractedwithin 24 h from collection. Physicochemical parameters including pH,temperature, and salinity were determined on-site using a CTD-typemulti-parameter probe (MS5, OTT Hydrolab, Germany). For qualitycontrol, field blanks, laboratory blanks, and laboratory-fortified blankswere sampled. Procedural standard calibration was performed usingspiked seawater calibration standards, which were treated exactly inthe same manner as samples. Internal standards were added to samplesand used as surrogates to monitor the reliability of complete analyticalprocedures. A set of at least seven laboratory spiked standard solutionswere analyzed to calculate the mean recovery (R) and the relativestandard deviation (RSD) (Table S1). The detection limits (DL) and thequantification limits (QL) for the analyzed chemicals and parametersand their estimation procedures are presented in supplementary in-formation (Table S1).

Sediments were collected using an Ekman grab sampler at 10sampling stations (8p, 11x, 12x, 13x, 14m, 15m, 17m, 19m, 22m, 24m)at depths indicated in supplementary information (Table S4). The col-lected sediments were stored at 4 °C away from sunlight.

2.3. Chemical standards

Analytical standards including THM calibration mix, halogenatedvolatile mix (containing HANs), and HAA esters calibration mix, 2,3-dibromopropionic acid solution, and the THA chloral hydrate werepurchased from Supelco (USA). The brominated THA tri-bromoacetaldehyde (97%) was purchased from Aldrich (UnitedKingdom) and was used to generate its hydrated form bromal hydrate(BH) in ultrapure water (Millipore, resistivity> 18MΩ cm). HPs 2-bromo-4-chlorophenol (98%) and 2,6-dibromophenol (99%) werepurchased from Alfa Aesar (Germany), and 2,4-dibromophenol (95%)and 2,4,6-tribromophenol (99%) were purchased from Sigma-Aldrich(USA). A standard stock solution of each compound was prepared inmethyl tert-butyl ether (MTBE, purity 99.8%) which was purchasedfrom Merck (Germany). L-ascorbic acid, crystalline, reagent grade waspurchased from Sigma (China). Sulfuric acid, analytical grade reagent,was purchased from Fisher Scientific (UK). For plotting the calibrationcurve, artificial seawater was spiked with the mother solutions at dif-ferent concentrations and the resulting solutions were treated accordingto the methods described hereby for samples. Artificial seawater (ASW)was prepared according to ASTM D1141-98 (2013).

2.4. Seawater sample preparation

For the analysis of THMs, HANs, THAs sample aliquots (50mL)

were first adjusted to a pH value ranging between 4.5 and 5.5 by addingsulfuric acid according to U.S. EPA Method 551 .1 with slight mod-ifications (Munch and Hautman, 1995). Internal standards were addedas surrogates to monitor the efficiency of sample treatment, and Na2SO4

(16 g) was then added. 1,2,3-trichloropropane was added as surrogatefor the analysis of THMs, HANs and BH. Samples were extracted byliquid-liquid extraction (LLE) by adding MTBE (5mL) and shakingmanually for 3min. Then, the organic phase was collected for analysis.For the analysis of HAAs, U.S. EPA Method 552.3 was used with slightmodifications (Domino et al., 2003). In brief, sample aliquots (40mL)were acidified to a pH < 1 by adding concentrated sulfuric acid andextracted with MTBE (4mL). 2,3-dibromopropionic acid was added tothe extracts as a surrogate. After LLE, the organic phase containing theHAAs was collected and transferred into 15mL vials to which acidifiedmethanol was added and placed in a water bath at 50 °C for 2 h forderivation (methylation). The vials were then cooled, and 4mL of sa-turated sodium bicarbonate solution were added before collecting theorganic phase containing the HAA esters in chromatographic vials. Forthe analysis of HPs, derivation (acetylation) and extraction by LLE wereconducted as described previously by Allonier et al. (1999a) with somemodifications. In brief, 50 mL samples were mixed manually during4min with 10 g of sodium carbonate and 5mL of acetic anhydride toderivatize the HPs. Samples were then extracted with 2.5 mL of MTBEcontaining the internal standard 2,4,6 trichlorophenol. The organicphases were then collected and dried with sodium sulfate before ana-lysis.

2.5. Analytical methods

Free residual oxidants and total residual oxidants were measured bythe colorimetric DPD method using a portable spectrophotometer(AQUALYTIC-AL 800, Germany). Bromide levels in water were mea-sured by an ICS-3000 Dionex ion chromatography system using a30mM NaOH eluent with a flow rate of 1.5mL/min at 30 °C. TotalOrganic Carbon (TOC) and Total Nitrogen (TN) were measured usinghigh temperature catalytic oxidation technique (Multi N/C 2100,Analytik Jena, Germany). Pre-treated samples were injected (50 μl) intothe furnace filled with a Pt preconditioned catalyst. Combustion wasrealized at 800 °C and combustion products were carried by high purityoxygen (Linde Gas) allowing detection of CO2 by non-dispersive in-frared (NDIR) and detection of NO by chemiluminescence (CLD).

Organic extracts containing CBPs were analyzed using a gas chro-matograph coupled to a 63Ni electron-capture detector (GC-ECD modelClarus 580, Perkin Elmer, Norwalk, CT, USA). An Elite 5MS capillarycolumn was used for the separation. Helium 5.0 was used as a carriergas at 1mL/min. Nitrogen was used as a make-up gas at 30mL/min.For the analysis of THMs, HANs, HPs and BH the temperature programwas as follows: initially 35 °C increasing to 145 °C at a rate of 10 °C/min, then at a rate of 20 °C/min up to 225 °C and finally at 10 °C/min to260 °C, held for 2min. For the analysis of HAAs, the temperature wasinitially set to 40 °C, then increased to 75 °C at a rate of 15 °C/min, thenincreased to 100 °C at 5 °C/min, and finally temperature reached 135 °Cat 10 °C/min which and held 2min. Analytes were qualified usingprocedural standard calibration. Calibrations were performed at con-centrations starting from as low as 10 ng/L up to 20 μg/L, depending onthe analytes (Table S1). At each concentration order, external calibra-tions were performed using a set of 7 standard solutions. The solutionswere prepared by adding aliquots of the standard stock solution in ar-tificial seawater, which was then treated exactly as a sample. Seawaterand purified water reagent blanks were included with each sequence Allanalytical method validation parameters are presented in Table S1.

2.6. Sediment sample preparation

Sediments were brought to room temperature. Extraneous materialwas removed prior to homogenization. Sediment samples were

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

3

homogenized in pre-cleaned collection jars by stirring vigorously withstainless steel spatulas. Dry weights were determined by placing samplealiquots in an oven at 105 °C and weighed at intervals of 24 h untilsuccessive weight differences became less than 4%. For the analysis ofHPs in sediments, the method was inspired from Lampi et al. (1992)with modifications. An aliquot (20 g) of wet sediments was weighed,and 50mL of 1M NaOH solution and internal standard (50 μL) of 2,4,6-trichlorophenol solution (1mg/L) were added. Then, 50 mL of hexanewere added to the sample and the whole kept in an ultrasonic bath for10min. The hexane phase was then discarded, and the aqueous phasewas collected and introduced into 65-mL glass vials with PTFE-linedscrew caps which were centrifuged at 2000 rpm for 10min. 40 mL ofthe resulting supernatant was then collected and was treated accordingto protocols used for the analysis of HPs in water samples describedabove.

3. Results and discussion

3.1. Physicochemical parameters

Water quality parameters can influence the formation and decay ofCBPs (Cimetiere et al., 2010; Hansen et al., 2012; Zhang et al., 2013). Atthe different sampling stations, physicochemical parameters includingwater temperature, pH, salinity, bromide concentration, TOC, TN, freeresidual oxidant, and total residual oxidant were determined in surfaceand bottom (or 7m deep) seawater (Table S2).

Very little pH variation was observed among the different samplingpoints located throughout the Gulf, with pH ranging between 8.13 and8.25 (Table S2). Water temperatures at the sampling stations close tothe shore were slightly higher than at offshore stations (Fig. 2), which islikely related to the discharge of heated seawater effluents by the in-dustrial sites. Additionally, at sampling stations near the outlets (9x,10x, 11x), water temperatures were slightly higher at the surface thanat the bottom (up to 18.9 °C and 19.8 °C, respectively) (Table S2). Atoffshore stations away from industrial outlets, water temperatures wereglobally homogeneous at the bottom and surface (around 14.5 °C),which is compatible with the agitated conditions that favor the mixingand homogenization of water bodies.

With regard to salinity, very minor variations were observed amongthe sampling stations located away from the shore (Fig. 1S). In contrast,near the shore, a marked influence of the Rhone River freshwater

intrusion appeared at nearby stations 2c, 8p and 5p, which had lowersalinities than the sampling stations located offshore. The samplingstation 2c located on the canal receiving water of the Rhône River, hadunusually high salinities (24.2 and 34.2 g/L on the surface and bottom,respectively). This salinity likely resulted from the vigorous mixing ofthe Rhône River freshwater with seawater entrained from the Gulf ofFos by the moderate to strong southeast winds registered on the day ofsampling. During the sampling campaign, sustained southeast wind(35–40 km/h on average) with gusts were reported (MétéoFrance, Is-tres). Like salinity, bromide concentrations varied very little amongstmost of the sampling stations (Fig. 3) and were generally high at the allof the sampling stations located in the Gulf (Table S2).

Total residual oxidant and free residual oxidant were not detected atany of the sampling stations. A number of factors could be at the originof this finding including low initial chlorination level, high organicmatter concentrations and temperatures (Brown et al., 2011). However,since the organic content in water (measured by TOC) and temperaturewere not unusually high (Table S2), the main factor is likely related tolow chlorination practiced by the industries during the sampling period.

3.2. Occurrence of chlorination byproducts (CBPs) in seawater

The levels of the CBPs measured at all the sampling stations arepresented in Table S3. Fig. 4 summarizes the sampling stations at whichCBPs were measured (whether on the surface or in the bottom) at levelshigher than QL. These CBPs were all brominated including bromoform,dibromoacetic acid (DBAA), dibromoacetonitrile (DBAN), tri-bromoacetic acid (TBAA), and BH. Low concentrations of bromo-chloroacetonitrile (BCAN) were also detected near some outlets (sam-pling stations 8p, 9x, and 10x) (Table S3). In previous studies ofchlorinated cooling waters, CBPs including THMs, HAAs, HANs, andHPs have been detected (Khalanski and Jenner, 2012), but to ourknowledge, this is the first study to report the occurrence of BH (hy-drated form of tribromoacetaldehyde), in the context of industrialseawater chlorination. Nevertheless, BH has been previously detectedin chlorinated seawater pools and in bromide-rich drinking waters(Krasner et al., 2006; Manasfi et al., 2017c).

The occurrence of predominantly brominated CBPs in chlorinatedseawater is in agreement with previous studies that explored the for-mation of CBPs in chlorinated seawater (Jenner et al., 1997; Allonieret al., 1999b; Boudjellaba et al., 2016). This speciation can be explained

Fig. 2. Variation of surface water temperature (in °C) at the different samplingstations.

Fig. 3. Variation of surface bromide concentrations (in mg/L) at the differentsampling stations. Triangles represent industrial effluent discharge points.

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

4

by the formation of bromine upon the addition of chlorine to seawater,which contains elevated bromide concentrations known to enhance theformation of brominated CBPs (Hua et al., 2006). In the presence ofappreciable amounts of bromide ions, chlorine oxidizes bromide ionsand forms hypobromous acid and hypobromite ions (Singer, 1999). Asan oxidant, bromine is stronger than chlorine and reacts 10 times fasterwith organic matter (Westerhoff et al., 2004). The reactions leading tothe formation of bromine in bromide-rich water such as seawater are asfollows (Heeb et al., 2014):

HOCl + Br− → HOBr + Cl - k1 = (1.55–6.84).103M−1 s−1 (1)

ClO− + Br− → BrO− + Cl− k2=9.10−4 M−1 s−1 (2)

Although HOBr and OBr− are the most abundant species involved inthese reactions, several bromine species can react with organic com-pounds present in seawater (Heeb et al., 2014; Manasfi et al., 2017a,b).Bromine species such as Br2, Br2O, BrOCl, and BrCl are less abundantbut more reactive than HOBr and OBr− and have been shown to con-tribute to the bromination of some organic compounds (Sivey et al.,2013). Table 1 shows the equilibrium concentrations of bromine specieswhen seawater containing 58mg/L bromide, at pH 8.2, is chlorinated at1mg/L active chlorine. These concentrations were estimated using

PHREEQC (Parkhurst and Appelo, 2013).The levels of CBPs were highest at the sampling stations located

near the outlets where chlorinated industrial effluents are released(stations 8p, 9x, 10x, 11x, and 13x) (Fig. 4). At these outlets, the con-centration of bromoform reached 1.95 μg/L at the surface and 2.36 μg/L at the bottom (Table S3). The slightly higher concentrations of bro-moform at the bottom compared to the surface may result from ac-celerated volatilization on the surface because of wind and the slightlyhigher temperatures. For the other CBPs, such distinctive discrepanciesbetween bottom and surface levels were not observed. DBAA, whichwas the second most abundant CBP, reached 1.40 μg/L at station 8p(surface). DBAN had a maximal concentration of 0.79 μg/L at station10x. At the sampling station 12x (in the vicinity of the outlet of LNGFos-Cavaou terminal), concentrations of CBPs were unexpectedly lowcompared to the other stations near effluent discharges. This aberrationis probably due to the dislocation of the sampling Niskin bottle underthe effect of wind during sampling. Strong southeast winds and currentsmade sampling very difficult at this station, located at about 150mfrom the shore. At sampling stations distant from the outlets of in-dustrial effluents such as 14m, 15m, 16m, 17m, only bromoform wasstill detected at levels above the QL (Fig. 5). For example, at stations14m, 15m, 16m, bromoform concentration was 0.16 μg/L. However,these bromoform concentrations at the offshore stations are still su-perior to typical background levels of bromoform found in seawater(emitted by marine algae) unexposed to chlorinated effluents. Bromo-form background levels have been estimated at 0.025 μg/L, and rarelyexceed 0.1 μg/L unless when extensive beds of macro-algae are present,which is not the case in the Gulf of Fos (Quack and Wallace, 2003).Thus, these above-background concentrations can be attributed to therelease of CBPs from industrial sites into the gulf.

In a previous survey, Jenner et al. (1997) investigated the occur-rence of CBPs in cooling water of several coastal power stations locatedin the UK, France, and the Netherlands. The analyzed CBPs includedTHMs, HANs, and HPs. Among these CBPs, bromoform and di-bromoacetonitrile (DBAN) were the main detected compounds in powerstations effluents at concentration that ranged from 0.72 to 29.20 μg/L,and from<0.1 to 3.15 μg/L, respectively (Jenner et al., 1997). More-over, Allonier et al. (1999b) measured THMs, HANs, HPs and HAAs in

Fig. 4. Concentrations of CBPs at sampling station where CBPs > QL were detected. B represent Bottom samples and S represents Surface samples. TBAA: tri-bromoacetic acid; BCAN: bromochloroacetonitrile; BH: bromal hydrate; DBAN: dibromoacetonitrile; DBAA: dibromoacetic acid. Errors bars are based on RSD% ofCBPs.

Table 1Molar concentrations of bromine species at equilibrium inseawater (Bromide 58mg/L; 20 g/L chloride, pH 8.2; chlori-nated with 1mg/L active chlorine), calculated usingPHREEQC (Parkhurst and Appelo, 2013).

Bromine Species Concentration (M)

HOBr 1.70× 10−5

OBr− 1.04× 10−5

Br2 7.62× 10−9

Br2Cl− 5.10× 10−9

Br2O 2.10× 10−9

BrCl 2.38× 10−10

Br3− 8.18× 10−11

BrCl2− 7.98× 10−11

BrOCl 5.33× 10−13

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

5

effluents of three French nuclear power stations after on-site chlorina-tion. The major compound formed in all power stations was bromoformat concentrations that ranged from 1.66 to 26.80 μg/L. DBAA andDBAN were also detected at levels ranging from 1.96 to 10.19 μg/L andfrom 0.94 to 3.61 μg/L, respectively. Accordingly, the levels of CBPsreported in these previous investigations of power station cooling watereffluents were higher than levels reported in the present study. How-ever, it should be noted that the maximal concentrations reported inthose previous studies of power station effluents were detected in thechlorinated effluents themselves, unlike in the current study where thecontamination of the receiving seawater at sampling stations dispersedinside the Gulf is explored, by measuring CBPs in the Gulf seawaterinstead of the effluents themselves. Furthermore, in these previousstudies total residual oxidant ranged from 0.3 to 1.5 mg/L (Jenneret al., 1997) and from 0.2 to 0.77 (Allonier et al., 1999b), suggestingstronger chlorination than what was practiced in the Gulf of Fos wheretotal residual oxidant levels were below QL (Table S2). A correlationbetween residual oxidant levels and CBP concentrations has been pre-viously suggested in previous studies of power station cooling effluents,with lowest CBP concentrations being detected at lowest total residualoxidant levels.

By comparing the findings of the present study conducted in springseason with the findings of the previous survey of Boudjellaba et al.(2016) conducted in the Gulf of Fos in winter and summer seasons, adiscrepancy appears between CBP levels in summer compared to thosein spring or winter which were quite similar. In the previous survey ofBoudjellaba et al. (2016), the used analytical methods had higher DLsand QLs resulting in undocumented levels for many of the analyzedCBPs at most of the sampling stations, as reported by authors. There-fore, the description of variation of CBP contamination among differentseasons can be chiefly performed by comparing the concentrations ofbromoform as it represented the most detected CBP in the past andpresent surveys. In Boudjellaba et al. (2016), bromoform concentrationsin surface seawater reached 2.51 μg/L and 7.55 μg/L in winter andsummer, respectively, with highest concentrations detected near out-lets. Other CBPs such as HAAs were only detected in summer withDBAA reaching 2.20 μg/L at the surface. In the present study, bromo-form concentration in surface seawater reached 1.95 μg/L, and DBAAconcentration reached 1.40 μg/L. Hence, it appears that the con-tamination of the Gulf of Fos by CBPs was highest in summer. Thisfinding is likely related to variations of industrial chlorination treat-ments as a function of water quality parameters especially temperature,which differ from one season to another. In summer, high water tem-perature (22–23 °C) would favor bacterial and algal development whichleads industries to increase chlorination to face the elevated biofouling

risks as compared to lower risks in colder seasons. In spring and winter,differences in bromoform concentration appear to be very minor. Itshould be noted that the water temperatures were also very similar:14.5 °C in spring campaign reported here, and 13 °C in winter campaignreported in Boudjellaba et al. (2016), which could result in similarchlorination practices performed by the industrial sites in these similarconditions.

Overall, the levels of CBPs measured in the Gulf were relatively low.Yet, although these low concentrations are not expected to constitute athreat to human health via direct exposure, concerns regarding poten-tial ecotoxic effects and the accumulation of CBPs in seafood (thus in-direct human exposure) cannot be dismissed, especially that CBPs aredischarged into seawater in a chronic manner. Additionally, the bro-minated speciation of the detected CBPs contributes to these concerns,since brominated CBPs are known to be far more toxic than theirchlorinated analogues (Richardson et al., 2007). Furthermore, the ex-istence of a potential cocktail effect where the overall toxicity of CBPs isaccentuated due to additive or synergistic interactions among differentcontaminants could also be an aggravating factor (Banerji et al., 2012).

3.3. Distribution of CBPs in seawater

Among the detected CBPs, bromoform was the most abundantspecies at most of the sampling stations (Fig. 6). This finding is inagreement with previous studies that investigated the occurrence ofCBPs in seawaters exposed to chlorinated industrial effluents (Jenneret al., 1997; Allonier et al., 1999b; Boudjellaba et al., 2016). DBAA wasthe second most abundant CBP at all the sampling sites except station11x where DBAN constituted the second most abundant species. Thisdiscrepancy can be attributed to the presence of nitrogen-containingorganic compounds near the industrial outlet 11x. Unlike DBAA, thetribrominated HAA, TBAA, was detected only at the sampling station8p. This finding may be explained by the lower stability of TBAA incomparison to DBAA (Zhang and Mineara, 2002; Manasfi et al., 2016).It has been reported that TBAA may decompose to form bromoform inaqueous solutions (Zhang and Mineara, 2002). HPs were not detected inseawater samples.

The distribution of CBPs observed here (THM most dominant fol-lowed by HAAs) is similar to the distribution observed in chlorinateddrinking water but distinct from that found in chlorinated swimmingpools, where HAAs represent the most dominant DBPs followed byTHMs (Chowdhury et al., 2014). The discrepancy between these dis-tributions is related to the nature of precursors leading to the formationof CBPs. In seawater and drinking water, NOM represent the main or-ganic precursors, while in swimming pools anthropogenic inputs are themain precursors. In a previous investigation (Kanan and Karanfil,2011), NOM were found to contribute mostly to the formation of THMs,while anthropogenic precursors (body fluids) contributed mostly to theformation of HAAs. Although this highlights the role of NOM in theformation of CBPs detected in the Gulf of Fos, the contribution of othertypes of precursors (e.g. synthetic organic compounds from industrialsites) can't be dismissed based on this observation.

3.4. Occurrence of brominated halophenols in sediments

Marine sediments were collected at the sampling stations (8p, 11x,12x, 13x, 14m, 15m, 17m, 19m, 22m, 24m). Two bromine-containingHPs were detected at the sampling stations 22m and 13x at con-centrations in the order of ng/g dry weight. At sampling station 22m, 2-chloro-4-bromophenol and 2,4,6-tribromophenol were detected at 1.8and 2.1 ng/g dry weight, respectively. At sampling station 13x, the twoHPs were detected at 0.3 and 1.5 ng/g dry weight, respectively.

Data about concentrations of halophenols in marine sediments inareas exposed to chlorinated effluents are very scarce. In a previousstudy, Sim et al. (2009) investigated the occurrence and distribution ofhalogenated phenols in sediments obtained from a marine environment

Fig. 5. Bromoform concentration (in μg/L) in bottom seawater at the samplingstations. Triangles represent industrial effluent discharge points.

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

6

near a nuclear power plant in South Korea. According to the findings ofthe latter study, bromophenols were found at higher concentrationsthan chlorophenols. The reported mean concentration of 2,4,6-tri-bromophenol reached 3.78 ng/g dry weight in spring. Furthermore,2,4,6-tribromophenol was previously detected at concentrations ran-ging from 2.80 to 10.39 μg/kg wet weight in the muscles of conger eelfish obtained from the Gulf of Fos (Boudjellaba et al., 2016). Based onthe geographical distribution of contaminated conger eel fish nearchlorination outlets in the latter study, authors suggested that thecontamination of fish by 2,4,6-tribromophenol resulted – at least par-tially – from chlorination water discharges. In the present study, thesampling stations where contaminated sediments were detected, werenot the ones located nearest to the chlorination outlets. Bromophenolscan be produced from both natural and anthropogenic sources (WHO,2005). Natural sources include marine organisms such as algae andbenthic animals, while anthropogenic sources include chlorination ofseawater and synthesis for use as intermediates in the production offlame retardants. In the analysis of HPs in seawater, all levels werebelow DL. However, very low levels below DL cannot be ruled out. Withthe effect of chronic presence, and a potential to sorb and accumulate insediments, this would lead to detectable and quantifiable levels in se-diments. Further investigations are necessary to discriminate whetherthe halophenols detected in sediments originate from the chlorinationof seawater or are produced naturally. Nevertheless, the occurrence of2,4,6-tribromophenol in the marine environment remains an issue ofconcern, since this compound has been reported to induce several sig-nificant adverse effects on fish populations (Deng et al., 2010).

4. Conclusions

This study measured the concentrations of CBPs in a semi-enclosedbay receiving industrial chlorinated seawater effluents. The highestlevels of CBPs including bromoform, DBAA, DBAN, and BH were de-tected near industrial outlets. To our knowledge, BH was reported herefor the first time in the context of industrial seawater chlorination.Other CBPs including TBAA and BCAN were detected at some samplingstations. At distant offshore sampling stations, only bromoform wasdetected at low levels. Furthermore, analysis of CBPs in sediments

showed the presence of two HPs (2-chloro-4-bromophenol and 2,4,6-tribromophenol) which were not detected in seawater samples.Although CBPs at these low concentrations aren't expected to constitutea threat to human health via direct exposure, concerns regarding eco-toxicity and indirect human exposure (via seafood) need to be ad-dressed. Despite low levels, a potential adverse effect to biota or ac-cumulation in seafood (thus indirect human exposure) can't be ruledout, especially that CBPs are being discharged into water in a chronicmanner. Therefore, to address these remaining concerns, suitable eco-toxicological assays of CBPs using adequate bioassays at en-vironmentally relevant concentrations and risk assessment studies arewarranted to assess the potential impact of the detected CBPs on en-vironmental health. It's also recommended to conduct an extensiveevaluation of contamination of marine organisms by CBPs in seawaterexposed to chlorinated effluents in future studies.

Conflicts of interest

Authors declare no conflicts of interests.

Acknowledgement

This work was funded by the French National Research Agency(ANR) within the project FOSSEA (ANR-16-CE34-0009). Authors ac-knowledge Météo-France for meteorological data through the agree-ment N° DIRSE/REC/16/02/0 and reviewers for helpful comments.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.ijheh.2018.06.008.

References

Allonier, A.S., Khalanski, M., Camel, V., Bermond, A., 1999a. Determination of dihaloa-cetonitriles and halophenols in chlorinated sea water. Talanta 50 (1), 227–236.

Allonier, A., Khalanski, M., Camel, V., Bermond, A., 1999b. Characterization of chlor-ination by-products in cooling effluents of coastal nuclear power stations. Mar. Pollut.Bull. 38, 1232–1241.

Fig. 6. Percentage distribution of the major CBPs in seawater near the chlorinated industrial outlets at the surface (upper figures) and at the bottom/depth (lowerfigures).

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

7

ASTM D1141-98, 2013. Standard Practice for the Preparation of Substitute Ocean Water.ASTM International.

Banerji, S., Werschkun, B., Höfer, T., 2012. Assessing the risk of ballast water treatment tohuman health. Regul. Toxicol. Pharmacol. 62 (3), 513–522.

Boudjellaba, D., Dron, J., Revenko, G., Demelas, C., Boudenne, J.L., 2016. Chlorinationby-product concentration levels in seawater and fish of an industrialised bay (Gulf ofFos, France) exposed to multiple chlorinated effluents. Sci. Total Environ. 541,391–399. https://doi.org/10.1016/j.scitotenv.2015.09.046.

Brown, D., Bridgeman, J., West, J.R., 2011. Predicting chlorine decay and THM formationin water supply systems. Rev. Environ. Sci. Biotechnol. 10 (1), 79–99.

Chowdhury, S., Alhooshani, K., Karanfil, T., 2014. Disinfection byproducts in swimmingpool: occurrences, implications and future needs. Water Res. 53, 68–109.

Cimetiere, N., Dossier-Berne, F., De Laat, J., 2010. Effect of some parameters on theformation of chloroform during chloramination of aqueous solutions of resorcinol.Water Res. 44, 4497–4504. https://doi.org/10.1016/J.WATRES.2010.06.010.

Deng, J., Liu, C., Yu, L., Zhou, B., 2010. Chronic exposure to environmental levels oftribromophenol impairs zebrafish reproduction. Toxicol. Appl. Pharmacol. 243 (1),87–95.

Domino, M.M., Pepich, B.V., Munch, D.J., Fair, P.S., Xie, Y., 2003. Method 552.3Determination of Haloacetic Acid and Dalapon in Drinking Water by Liquid-liquidMicroextraction, Derivatization, and Gas Chromatography with Electron CaptureDetection. Environ. Prot. Agency, Cincinnati, OH.

Hansen, K.M.S., Willach, S., Antoniou, M.G., Mosbæk, H., Albrechtsen, H.-J., Andersen,H.R., 2012. Effect of pH on the formation of disinfection byproducts in swimmingpool water – is less THM better? Water Res. 46, 6399–6409. https://doi.org/10.1016/j.watres.2012.09.008.

Heeb, M.B., Criquet, J., Zimmermann-Steffens, S.G., Von Gunten, U., 2014. Oxidativetreatment of bromide-containing waters: formation of bromine and its reactions withinorganic and organic compounds - a critical review. Water Res. 48, 15–42. https://doi.org/10.1016/j.watres.2013.08.030.

Hua, G., Reckhow, D.A., Kim, J., 2006. Effect of bromide and iodide ions on the formationand speciation of disinfection byproducts during chlorination. Environ. Sci. Technol.40 (9), 3050–3056.

Inkscape, 2015. http://www.inkscape.org.Jenner, H.A., Taylor, C.J.L., Van Donk, M., Khalanski, M., 1997. Chlorination by-products

in chlorinated cooling water of some European coastal power stations. Mar. Environ.Res. 43, 279–293.

Kanan, A., Karanfil, T., 2011. Formation of disinfection by-products in indoor swimmingpool water: the contribution from filling water natural organic matter and swimmerbody fluids. Water Res. 45 (2), 926–932.

Khalanski, M., Jenner, H.A., 2012. Chlorination chemistry and ecotoxicology of themarine cooling water systems. In: Rajagopal, S., Jenner, H.A., Venugopalan, V.P.(Eds.), Operational and Environmental Consequences of Large Industrial CoolingWater Systems. Springer Science Publisher, pp. 183–226.

Krasner, S.W., Weinberg, H.S., Richardson, S.D., Pastor, S.J., Chinn, R., Sclimenti, M.J.,Onstad, G.D., Thruston, A.D., 2006. Occurrence of a new generation of disinfectionbyproducts. Environ. Sci. Technol. 40 (23), 7175–7185.

Lampi, P., Tolonen, K., Vartlalnen, T., Tuomisto, J., 1992. Chlorophenols in lake bottomsediments: a retrospective study of drinking water contamination. Chemosphere 24,1805–1824.

Ma, Z., Gao, K., Li, W., Xu, Z., Lin, H., Zheng, Y., 2011. Impacts of chlorination and heatshocks on growth, pigments and photosynthesis of Phaeodactylum tricornutum(Bacillariophyceae). J. Exp. Mar. Biol. Ecol. 397 (2), 214–219.

Manasfi, T., De Méo, M., Coulomb, B., Di Giorgio, C., Boudenne, J.L., 2016. Identificationof disinfection by-products in freshwater and seawater swimming pools and eva-luation of genotoxicity. Environ. Int. 88, 94–102. https://doi.org/10.1016/j.envint.2015.12.028.

Manasfi, T., Coulomb, B., Ravier, S., Boudenne, J.-L., 2017a. Degradation of organic UVfilters in chlorinated seawater swimming pools: transformation pathways and bro-moform formation. Environ. Sci. Technol. 51 (23), 13580–13591. https://doi.org/10.1021/acs.est.7b02624.

Manasfi, T., De Méo, M., Di Giorgio, C., Coulomb, B., Boudenne, J.L., 2017b. Assessing

the genotoxicity of two commonly occurring byproducts of water disinfection :chloral hydrate and bromal hydrate. Mutat. Res. Genet. Toxicol. Environ. Mutagen813, 37–44. https://doi.org/10.1016/j.mrgentox.2016.11.009.

Manasfi, T., Temime-Roussel, B., Coulomb, B., Vassalo, L., Boudenne, J.L., 2017c.Occurrence of brominated disinfection byproducts in the air and water of chlorinatedseawater swimming pools. Int. J. Hyg Environ. Health 220 (3), 583–590.

MeteoFrance. http://www.meteofrance.fr/publications/glossaire/152770-mistral.Munch, D.J., Hautman, D.P., 1995. Method 551.1. Determination of Chlorination

Disinfection by Products, Chlorinated Solvents, and Halogenated Pesticides/Herbicides in Drinking Water by Liquid-liquid Extraction and Gas Chromatographywith Electron Capture Detection. Methods for the Determination of OrganicCompounds in Drinking Water, Supplement III, EPA-600/R-95/131. Cincinnati, OH.U.S. Environmental Protection Agency, Cincinnati, OH.

Parkhurst, D.L., Appelo, C.A.J., 2013. Description of Input and Examples for PHREEQCVersion 3: a Computer Program for Speciation, Batch-reaction, One-dimensionalTransport, and Inverse Geochemical Calculations (No. 6–A43). US Geological Survey.

Pignata, C., Fea, E., Rovere, R., Degan, R., Lorenzi, E., de Ceglia, M., Schilirò, T., Gilli, G.,2012. Chlorination in a wastewater treatment plant: acute toxicity effects of the ef-fluent and of the recipient water body. Environ. Monit. Assess. 184 (4), 2091–2103.

Quack, B., Wallace, D.W.R., 2003. Air-sea flux of bromoform: controls, rates, and im-plications. Global Biogeochem. Cycles 17 (1), 1023.

R Core Team, 2015. R: a Language and Environment for Statistical Computing. RFoundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/.

Richardson, S.D., Plewa, M.J., Wagner, E.D., Schoeny, R., DeMarini, D.M., 2007.Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfectionby-products in drinking water: a review and roadmap for research. Mutat. Res. Rev.Mutat. Res. 636 (1), 178–242.

Sim, W.J., Lee, S.H., Lee, I.S., Choi, S.D., Oh, J.E., 2009. Distribution and formation ofchlorophenols and bromophenols in marine and riverine environments. Chemosphere77 (4), 552–558.

Singer, P.C., 1999. Humic substances as precursors for potentially harmful disinfectionby- products. Water Sci. Technol. 40, 25–30. https://doi.org/10.1016/S0273-1223(99)00636-8.

Sivey, J.D., Arey, J.S., Tentscher, P.R., Roberts, A.L., 2013. Reactivity of BrCl, Br2, BrOCl,Br2O, and HOBr toward dimethenamid in solutions of bromide + aqueous freechlorine. Environ. Sci. Technol. 47, 1330–1338. https://doi.org/10.1021/es302730h.

Stewart, M.E., Blogoslawski, W.J., Hsu, R.Y., Helz, G.R., 1979. By-products of oxidativebiocides: toxicity to oyster larvae. Mar. Pollut. Bull. 10 (6), 166–169.

Taylor, C.J.L., 2006. The effects of biological fouling control at coastal and estuarinepower stations. Mar. Pollut. Bull. 53 (1–4), 30–48.

Ulses, C., Grenz, C., Marsaleix, P., Schaaff, E., Estournel, C., Meulé, S., Pinazo, C., 2005.Circulation in a semienclosed bay under influence of strong freshwater input. J. Mar.Syst. 56 (1–2), 113–132.

Werschkun, B., Banerji, S., Basurko, O.C., David, M., Fuhr, F., Gollasch, S., Grummt, T.,Haarich, M., Jha, A.N., Kacan, S., Kehrer, A., 2014. Emerging risks from ballast watertreatment: the run-up to the international ballast water management convention.Chemosphere 112, 256–266.

Westerhoff, P., Chao, P., Mash, H., 2004. Reactivity of natural organic matter withaqueous chlorine and bromine. Water Res. 38, 1502–1513. https://doi.org/10.1016/j.watres.2003.12.014.

World Health Organization (WHO), 2005. 2,4,6-Tribromophenol and Other SimpleBrominated Phenols. Concise International Chemical Assessment Document (CICAD)66. World Health Organization, Geneva.

Yang, M., Zhang, X., 2013. Comparative developmental toxicity of new aromatic halo-genated DBPs in a chlorinated saline sewage effluent to the marine polychaetePlatynereis dumerilii. Environ. Sci. Technol. 47 (19), 10868–10876.

Zhang, X., Minear, R. a., 2002. Decomposition of trihaloacetic acids and formation of thecorresponding trihalomethanes in drinking water. Water Res. 36, 3665–3673.https://doi.org/10.1016/S0043-1354(02)00072-6.

Zhang, X.L., Yang, H.W., Wang, X.M., Fu, J., Xie, Y.F., 2013. Formation of disinfection by-products: effect of temperature and kinetic modeling. Chemosphere 90 (2), 634–639.

T. Manasfi et al. International Journal of Hygiene and Environmental Health 222 (2019) 1–8

8